Confusing dac and amp

[Bliman]

Hi,
I am someone who takes a long time before I buy something.
I have a Beyerdynamic DT250 -250 Ohm. And for the moment it is plugged in a Grace Design m902. But the Grace Design will go to my main system again very shortly, so it will not be used with my computer anymore.
Now I have compared the Grace design with the output of my computer and I hear a difference between the two. The computer is more dense and more thinner in sound. But the difference is not much.
Now I don't know if it is wise to buy a amp and dac.
There are also some things that I don't understand. The Atom amp is being described as one of the best measured amps here.
But I saw this site and head amp and it looks really nice but when I look at the specs it looks pretty bad.
https://rupertneve.com/products/rnhp-headphone-amplifier/#tech-specs
It get's glowing reviews and is also from a brand that is very serious.
How can this be? With that I am saying is, are measurements the be it end all?
Or is this head amp automatically not as good?
I am also having difficulty determining how much money I should give for an amp and dac. If you know that my headphone costs 137€.
Is it a waste of money?
And how much should I allocate to both?
Should I give much less to a dac because the technology goes faster on this side.
So many questions
Thanks for reading.

 

[MZKM]

But I saw this site and head amp and it looks really nice but when I look at the specs it looks pretty bad.

What bad spec(s) are you referring to?

are measurements the be it end all?

Measurements do not have a strong correlation to sighted listening impressions. I’ve heard speakers that I know should/do measure poorly, but sounded fine to me. If a product gives off the impression that it should sound good (looks, price, brand, reviews, etc.), then you likely will have a positive bias, whereas the opposite is also true.

See this small segment from a talk Toole gave:

 

[Bliman]

Noise

Measured with typical headphones, Z=44 Ω, BW 22Hz – 22kHz
"A" XLR Input: -101.9dBV,
"B" RCA Input: -100.9dBV
"C" 3.5mm Input: -88.8dBV
But bare in mind that my technical knowledge is close to zero. Certainly with the people here.

 

[MZKM]

Noise

Measured with typical headphones, Z=44 Ω, BW 22Hz – 22kHz
"A" XLR Input: -101.9dBV,
"B" RCA Input: -100.9dBV
"C" 3.5mm Input: -88.8dBV
But bare in mind that my technical knowledge is close to zero. Certainly with the people here.

If you think noise that’s 100dB lower than your signal is poor, then you have ridiculously high standards.

 

[Bliman]

If you think noise that’s 100dB lower than your signal is poor, then you have ridiculously high standards.

ok I thought it was pretty poor. Isn't the atom 113-114 dBV?
I am a great noobie here of course. I just want to make the best decisions.
And if someone or you have answers to my other questions. That would be great

 

[MZKM]

ok I thought it was pretty poor. Isn't the atom 113-114 dBV?
I am a great noobie here of course. I just want to make the best decisions.
And if someone or you have answers to my other questions. That would be great

Yes, the Atom is that high, hence it treated as the best headphone amp under $300.

However, you really should familiarize yourself with how huge of difference 100dB is. Going from a 1W amp to a 100W amp is only a 20dB difference.

 

[M00ndancer]

I am also having difficulty determining how much money I should give for an amp and dac. If you know that my headphone costs 137€.

If you want a replacement for the Grace for your computer you have a couple of options.
First, get a Atom headphone amp and see if that's enough of an improvement. (please do a real blind test)
Second, if its not get a Topping D10, D30 or a khadas board.
If you really want to save money you can get a HAD-1 like me that's both a DAC/AMP like the Grace m902.
Start with what type of features you need, find all the DACs, AMPs or DAC/AMPs that fulfills the features.
Then check the master test list here https://www.audiosciencereview.com/...ad-distortion-comparison-graph-for-dacs.4814/

The bottom line is that there are excellent cheap choices, it all comes down to what features you want/need.

 

[Bliman]

dB is log is I remember correctly. But then again the Atom scores better then the more expensive Rupert Neve.
There is a mention about the placebo effect here and I agree that is a big one, not only in audio but in all fields.
But then again you look at the Rupert Neve RNHP and it get's good reviews and it is a pro company.
I just don't get it. The Atom has basically perfect scores. Then why are studio's not using them?
Is it all a moneygrab?
I liked the Rupert Neve (I haven't tested it) because it comes from a pro audio company and pro companies usely bill their equipment fairer then the audiophile companies.
I read many reviews here and it is often that it is stated with many measurements that it is so low that you can't hear it.
Then the Atom can replace all the amps in regard to sound quality even those that measure great but cost many thousands.
It is all so confusing to me.
Are people saying here that if you take for example two good dacs or two good amps and blindly listen to them that you would hear no difference?

 

[RayDunzl]

It is all so confusing to me.

Did you visit the Shoutometer?

 

[MZKM]

Are people saying here that if you take for example two good dacs or two good amps and blindly listen to them that you would hear no difference?

If distortion/noise/“colorations”/etc. are all below audible levels, then yes.

Take note though that we have to look at the performance of the system, and not a single component to talk about this. 1% THD in the treble is about the treshold for audibility while music is playing, but that doesn’t mean you can have 1% THD for your DAC, preamp, DSP device, amp, etc., as THD, and other distortions, stack/compound; not to mention the speakers. Some speakers already have enough distortion to be audible on their own, so any gear you use, regardless of performance, is only going to increase this already audible distortion.

 

[Bliman]

Did you visit the Shoutometer?

Thank you very much. Very enlightening. Thanks for teaching me.

 

[Bliman]

If distortion/noise/“colorations”/etc. are all below audible levels, then yes.

Take note though that we have to look at the performance of the system, and not a single component to talk about this. 0.1% THD in the treble is about the treshold for audibility while music is playing, but that doesn’t mean you can have 0.1% THD for your DAC, preamp, DSP device, amp, etc., as THD, and other distortions, stack/compound; not to mention the speakers. Some speakers already have enough distortion to be audible on their own, so any gear you use, regardless of performance, is only going to increase this already audible distortion.

So all these distortions multiply or add up? Or is it as good as the one with the worst THD?
I see at the specs of Beyerdynamic DT 250 that it has a THD of <0.2. I guess that is a lot.
These are the specs of the Grace design m902
Thd+n+10dBu out, 50 Ohm load, SMPTE 4:1<0.01%Headphone +10dBu out, 50 Ohm load<0.008%Line Out +10dBu out<0.02%
Am I correct in thinking that the atom will not be any better then this then?
And that there will be 0 % change in sound if I swap them? Or is this incorrect?
Also thank you for helping

 

[RayDunzl]

Thank you very much. Very enlightening. Thanks for teaching me.

I don't claim any accuracy to it, but it might provide some perspective to the negative numbers.

 

[MZKM]

So all these distortions multiply or add up? Or is it as good as the one with the worst THD?

Stacks.

20*log10(10^(x/20) + 10^(x/20) + ...)

If you have 4 components (speakers, DAC, preamp, power amp), and they all have 1% THD (-40dB), that would be:

20*log10(4•10^(-40/20)) = ~ -28dB (~4% THD)

EDIT: This is for correlated “distortion” as @RayDunzl stated; channel separation (crosstalk) would be an example, but THD across components likely would be uncorrelated.

 

[RayDunzl]

So all these distortions multiply or add up?

Amplification is a multiplication of voltage - the source noise and the signal would be amplified (multiplied), but to the same ratio, so the difference would remain the same - ignoring the noise of the amplifier itself, which would be added.

If added together, the increase requires examination of whether the items being added a coherent/correlated (in step with each other) or incoherent/uncorrelated (unaligned).

Aligned noises of the same level could add together to increase as much as 6dB, unaligned as much as 3dB.

If the levels are different, then the result is much closer to just the louder one.

Decibel addition (coherent/correlated):
-100dB + -100B = -93.979dB --- 0.001% to 0.0020001%
-100dB + -115dB = -98.578dB --- 0.001% to 0.0011779%

Decibel addition (incoherent/uncorrelated):
-100dB + -100B = -96.99dB --- 0.001% to 0.0014142%
-100dB + -115dB = -99.865dB --- 0.001% to 0.0010157%

 

[Bliman]

Amplification is a multiplication of voltage - the source noise and the signal would be amplified (multiplied), but to the same ratio, so the difference would remain the same - ignoring the noise of the amplifier itself, which would be added.

If added together, the increase requires examination of whether the items being added a coherent/correlated (in step with each other) or incoherent/uncorrelated (unaligned).

Aligned noises of the same level could add together to increase as much as 6dB, unaligned as much as 3dB.

If the levels are different, then the result is much closer to just the louder one.

Decibel addition (coherent/correlated):
-100dB + -100B = -93.979dB
-100dB + -115dB = -98.578dB

Decibel addition (incoherent/uncorrelated):
-100dB + -100B = -96.99dB
-100dB + -115dB = -99.865dB

Thanks. But to be honest this is already a bit above my head.
Do you guys are engineers?
Is it good if I learn first the Maxwell's equations? And is there a book that gently walks me in the matter?
Also is it correct in saying that the Atom with a good measuring dac like the el dac would sound no different then the Grace design M902?

 

[RayDunzl]

Calculations thanks to http://www.sengpielaudio.com/Calculations03.htm

 

[MZKM]

Thanks. But to be honest this is already a bit above my head.
Do you guys are engineers?
Is it good if I learn first the Maxwell's equations? And is there a book that gently walks me in the matter?
Also is it correct in saying that the Atom with a good measuring dac like the el dac would sound no different then the Grace design M902?

Pretty much every DAC whose SINAD falls in tier-3 or better on the chart will be transparent, assuming the performance of the other components in your chain aren’t bad enough to cause the performance of the system to have audible distortions/“colorations”/noise/etc.

 

[Bliman]

Thank you both.

 

[thunderchicken]

Someone above mentioned it already- your best bet for a completely transparent, neutral DAC/amp combo is going to be a Khadas tone board and a JDS Atom. At that point, you're at state of the art performance for 300 euros or so. You'll hear your headphones' sound signature and your music and not much else. If that's your thing, it's an easy decision. If you have more budget, there are a couple of other options- like the freshly reviewed Gustard DAC and the Neurochrome amplifier, which, depending on your headphones and general setup, might not get you any more performance than the budget options.